Endosomes moved toward the cell center in the control cells, but in LatB-treated cells few movements toward the perinuclear region were observed despite frequent random movements. The tracking analysis clearly indicated that actin polymerization was required for directional movement toward the perinuclear region. Next, to observe more SB 332235 directly the endosomal movements induced by actin polymerization, we used both LatB and nocodazole, followed by removal of only LatB to induce actin polymerization. When cells were treated with both LatB and nocodazole, enlarged EEs containing several clusters of EGF were observed. However, after removal of LatB, the EGF clusters spread quickly and had both tubular and vacuolar domains. As time progressed, EGF-containing vacuolar domains dispersed further and Tfn gradually disappeared. To quantify this dispersion, we measured the area of endosomes. The EE area was significantly increased at 5 min after the LatB washout. These results indicate that actin polymerization is required for both inhibition of homotypic fusion of endosomes through microtubule-independent movements and transport from EEs. The recruitment of actin filaments to EEs has been observed previously. Here, we demonstrated that inhibition of actin dynamics led to the formation of enlarged EEs and impaired transport from EEs. Simultaneous application of LatB and LY significantly inhibited Tfn recycling compared with individual LatB and LY applications. These results suggest that LatB and LY act in distinct pathways, and that actin might be involved in the EE-to-RE pathway, which is independent of PI3K. In fact, LatB treatment resulted in reduction in tubule formation from EEs. This leads to the major question, what is the role of actin in transport from EEs? Recently, it has been shown that actin dynamics induce scission of membrane tubules. Other researchers have also suggested that actin dynamics play a role in membrane scission. These experiments focused on the internalization steps at PM, but this actin-induced scission may also apply to EEs. SNX4 has been shown to be a candidate SB 258585 hydrochloride factor driving membrane tubulation in the EE-to-RE pathway, and may contribute to membrane tubulation and scission together with actin dynamics. Another study also reported that myosin VI and its interacting protein lemur tyrosine kinase 2 siRNAs led to swollen, enlarged EEs and reduced EHD3- containing tubule formation. These results suggest that actin motor proteins also participate in the EE-to-RE pathway. Indeed, at the trans-Golgi network, GOLPH3 bridges phosphatidylinositol and actomyosin to promote efficient tubulation and vesicle formation.